Analytics is the process of using data to inform managerial decision-making.

• What are the key managerial decisions in marketing?
• What is the difference between inform and drive?
• What are some examples of marketing analytics in practice?

To accomplish this, in this course we will be learning to analyze data in R.

“Good inspiration is based upon good information.” -President Russell M. Nelson (April 2018)

• Bad information can limit inspiration.
• Scientific and spiritual learning run in tandem.
• We need to be willing to learn from evidence.
• Admitting uncertainty isn’t a bad thing.

Nothing. (Well, nothing about coding.)

• We focus on secondary data instead of primary data.
• Secondary data is messy and comes from many different sources.
• This course is a data analysis deep dive: more code, more details, more technical.

• This a deeper dive into data analysis.
• Class is interactive.
• Many students find they learn more.

R is a free, open-source programming language for statistical computing, analysis, and data science.

• Largest repository of established and new statistical techniques.
• Friendly for non-programmers.
• A very active and helpful community.
• What you can do with R is in high demand.

• Visualize, wrangle, and summarize data.
• Acquire data from a variety of sources (e.g., databases, web scraping).
• Produce reports and interactive dashboards.
• Implement a variety of inferential and predictive models.

• You’ll find that no single programming language is the best at everything.
• I switch between R and Python for different tasks for the same project.
• R is the most accessible, general-purpose data analysis tool I can teach you.

Python is also a general-purpose data analysis tool, it just isn’t as accessible as R is. However, many jobs require you know Python.

• Programming techniques and experience learned in R will make learning additional languages much easier.
• The BYU Statistics department has excellent courses to teach you Python.
• If you are interested, I can provide access to resources for self-driven study.

1. Seek learning by study and faith (D&C 109:7).
2. Prepare for class by previewing material and coming with questions.
3. Actively code, take notes, and ask questions during class.
4. Practice coding by completing exercises, referencing supplementary material as needed.
5. Review exercise solutions and note where and why your work differs.
6. Use the quizzes to gauge how well the material is understood.
7. Work with classmates and utilize office hours.
8. Download and organize all course materials, notes, and code.

1. Learn by doing: Code in class and complete exercises.
2. Pay careful attention to details. READ SLOWLY.
3. Don’t code from scratch. Start with previous work and solutions.
4. Look at and emulate good code.
5. Literally sketch what transformed data should look like in the end.

• Email me: cameron.bale@byu.edu
• Office Hours: By appointment
• Contact the TA: esbudge@student.byu.edu, Phone: 636-226-5419 (for emergencies only, or if email unanswered after 72 hours)
• Go to TA office hours: 4pm - 6pm Mondays and Wednesdays. Zoom link.
• Reference slides, class notes, and supplementary material first
• Learn to use GitHub Copilot (eventually)

This class provides three detailed demonstrations of how marketing analytics is used in practice. Each unit is motivated by a case study with a corresponding data set.

• We use the case study data in class for the entire unit.
• You’ll use the case study data for the exercises.
• Each unit culminates in a project that finalizes answers for the case study.

• The beginning of the semester can have a steep learning curve. The first unit will likely be the most time consuming.
• Project weeks can be more intense than other weeks.
• Otherwise, expect the usual two hours a week for every hour spent in class (your mileage may vary).

“There is no way to go from knowing nothing about a subject to knowing something about a subject without going through a period of much frustration and suckiness. Push through. You’ll suck less.” -Hadley Wickham

• Don’t be afraid of this learning pit.
• You aren’t alone.
• If you commit to learning, I’ll gladly walk with you.
• Learn something hard now while you have help.
• Please be patient with yourself, me, and others.

Justin Collings, Academic Vice President (Remember the Alamo Bowl):

‘’Ultimately, I suggest, generosity is about recognizing and honoring a common familial bond with our fellow children of God. It is by virtue of that recognition that ’all relationships within the BYU community [can] reflect devout love of God and a loving, genuine concern for the welfare of our neighbor.’’’

To inform managerial decision-making there needs to be a decision to begin with. Having a well-defined managerial decision is what separates data analytics from data mining.

What data is needed should be evaluated with the managerial decision in mind and not the other way around.

While primary data is gathered specifically to serve the research objectives at hand, secondary data was gathered for another purpose. Consult the original and most current source whenever possible and understand its context (i.e., use the data dictionary).

• Who collected it?
• Why was it collected?
• When was it collected?
• What was collected?
• How was it collected?

Acquiring secondary data may require interfacing with databases, using APIs, scraping the Web, etc.

Secondary data can be messy. Data wrangling includes whatever tidying, cleaning, mutating, munging, selecting, transforming, renaming, fusing, or filtering is needed to get the data into the needed form to summarize and model.

This can be tedious.

Summarizing data is initially about discovery. It includes computing statistics (i.e., numerical summaries) and data visualization (i.e., graphical summaries).

• Summarizing data is closely tied with data wrangling.
• Summarizing data is often not an end in itself.

Models extract information from the data to inform our managerial decision.

• In order to inform the marketing mix, the models we use are often inferential.
• Some managerial decisions only rely on prediction.

Effectively communicating marketing insights brings us full circle and highlights the necessity of domain expertise.

The analyst needs to interpret results in a way that clearly informs the managerial decision. You may hear this referred to as “storytelling.”

Summary

• Defined marketing analytics.
• Discussed the FAQ.
• Walked through the marketing analytics process.

Next Time

• Getting started with R.
• Transforming data with {dplyr}.

Artwork by @allison_horst

1. Read the syllabus.
2. Sign up for Posit Cloud and join the course here.
3. Email me with questions or concerns you haven’t had answered and/or what topics you’re most excited to cover.
4. Read the case and write how you might meet the expectations (no more than one page).
5. Submit your response as a Word document on Canvas by the beginning of class Tuesday and be prepared to share with the class.